In this dissertation, I reviewed the functional aspects of venom from a behavioral perspective to develop a comprehensive understanding of the behavioral ecology of venom. To explore the behavioral and physiological ecology of venom expenditure by scorpions, I conducted four studies of the medically significant Buthid scorpion Parabuthus transvaalicus. The first two experiments revealed that scorpions regulate their venom expenditure during defensive stinging and squirting in the most complex manner yet described for any venomous organism. When stinging, these scorpions can choose between delivering a dry or wet sting. Should they deliver a wet sting, they meter the volume of venom injected, delivering more during high-threat than low-threat conditions. By metering volume, they also vary the composition of the venom, injecting either clear (potassium-rich) "prevenom" or milky (protein-rich) "venom." The milky venom is ejected only after the limited quantity of prevenom has been exhausted, usually after one or several low-volume stings. These scorpions also possess the capacity to squirt venom when grasped by the tail. Experimental evaluation of the stimuli eliciting squirts and videotape analysis of the squirt trajectory suggests that squirting serves an antipredatory function. In contrast to stinging, scorpions always eject milky venom when squirting. Collectively, these studies support the venom-metering hypothesis, which proposes that animals make cognitive decisions about their venom use. Two additional studies confirm the high metabolic cost of venom replacement. When scorpion venom glands were emptied, there was a significant increase in oxygen consumption during the subsequent 72 h, suggesting that venom resynthesis is an expensive metabolic investment. However, the regenerated venom had considerably lower protein concentration than the initial venom. A longer-term (192 h) study of venom replenishment in milked scorpions provided further insight. Lethality tests in crickets indicated that killing effectiveness of the replenished venom had returned by day 4. However, the gradual accumulation of major peptides in the reconstituted venom, detected by MALDI-TOF, and irregular spikes in oxygen consumption suggested that regeneration of different venom components was asynchronous during the 8-day period. These studies support the view that venom is a limited commodity and, therefore, should be used judiciously by scorpions.